If you asked us for an epic tale of designing and building under a deadline, one of the last places we would think to look is a MrBeast video. Yet here we are, thanks in no small part to the epic skills of one [William Osman].
What do you do when a major YouTube celebrity asks you to handle a project with an impossible deadline? If you’re [William], you say “heck yeah” and figure out the details later. In this case, it was famed YouTuber [Jimmy Donaldson], aka MrBeast, who was planning his own version of Squid Game. In this version, no one dies, but a few players do walk away with a lot of cash.
The premise is simple — “kill” people with a motion-sensing gun turret, just like the one in the show. The problem is that the show had all the tools of movie magic – multiple takes, video editing, you name it. [William] was tasked with handling a live event, with 456 real people, and no do-overs. Oh, and the whole thing had to be ready in 3 weeks.
The kills had to be pretty obvious too – we’re talking simulated blood squirting everywhere. So [William] decided to build his own version of a blood squib – the device Hollywood has used for decades to simulate bullet wounds. Initial work with pneumatic systems proved to be impractical. That’s when he put on his manager hat — and hired people to solve the problem for him. You might recognize a few of them — [Allen Pan] makes an appearance, as well as chemical genius [NileRed]. Even [TheBackYardScientist] shows up.
The video documents [William]’s journey, getting 500 copies of a board built and delivered on deadline. As such, there isn’t a ton of detail about the internal workings of the system. A pair of AA batteries feed into a boost converter, which powers an ESP8266 inside an ESP-WROOM-02 module. The ESP drives a few LEDs and a MOSFET. The MOSFET is connected to the star of the show – an MGJ firewire initiator – think of it as a model rocket igniter on steroids. The initiator hides behind a bag of YouTube-friendly yellow “blood”. When the system is commanded to kill, the initiator pops the bag, spraying blood everywhere.
Doing this for one device isn’t so bad, but we’re following Squid Game rules – which means 456 competitors. Further, there were 100 iPhones loaded with a custom kill app for the workers. Add a central server into the mix, and you’ve got 557 devices in close quarters basting on 2.4 GHz and 5.8 GHz. Did we mention that [William] had never done a test with more than a handful of devices?
Want to find out what happens? Check out the video below!
35 thoughts on “Engineering On A Deadline For Squid Game”
the video basically boils down to digikey sending a reel with every component placed in individual baggies.
Actually, I haven’t even understood the text article… (though I admitedly rather ‘skimmed’ it)
Unfortunately, so do I. I don’t know if I’ve become too impatient for such long videos or too unfocused. After 20 minutes, my brain shut down.
I meann there’s also the custom speed control they recently added to the YouTube video settings menu.
With limited time and young kids I can’t stand taking the time to read an article (probably with 2 or 3 interruptions and stop/starts) only to find out it was just burying the lead for a 20 minutes (or even a 30 second ticktock) video. Then I have to hunt down headphones or subject the whole family to the video and hope I can have a couple of uninterrupted minutes to watch what I could have read in 30 seconds.
I wouldnt call William Osmands skills “epic”, but that’s part of what makes his videos so funny. Very relatable mishaps too as an EE myself.
He also intentionally hams things up for the camera. Early on he tried to do more serious maker things and did not get much traction until he started making goofy ideas.
I understand why creators do this, and it is their choice to make, they need to eat. But the moment this happens, I lose all interest. I’m all for fun and bodging and doing stupid shit, but the moment it turns into purposefully screwing up for ‘content’ it just isn’t interesting anymore.
Like that time he made that haphazard X-ray unit and got salty when literally everyone told him it wasn’t safe, for so many different reasons. Not only was he scattering X-rays everywhere, but the potential for receiving a career limiting shock was real.
My favourite thing about this video is that he was saved by a small mom and pop pcb assembly joint. Large companies might have the capacity but as he said in the video you can’t pay them to care.
What? Jlpcb, pcbway, etc could have easily done this.
omg youtube cant ahow fake blood…. unless its from hollywood. i hate youtube.
Well, you can… lot’s of homemade action movies on there. But if you want to target certain audiences, or monetize it, you might wanna avoid such things
Ok, so someone spent $3.5 million dollars on the show, but only allowed 3 weeks for the entire design and build of one of the core elements?
welcome to dystopia. where real work is no longer valued and the show must go on.
It’s not just this project. The whole worlds supply chains rely on Just In-time Delivery including food supply chains.
We saw minor disturbances and product non-availabilities at the early stages of the pandemic. I fear what might happen if the world not has global political disruption event.
I think you are missing the point, the whole project was a rush job. They wanted to get the video out while squid games was still relevant. That’s why they they spent so much money to get it done quickly.
From what I’ve heard this is not unusual (uncommon yes, but not wildly unexpected) for TV/movie stuff. Especially if the production crew are not technical / not familiar with why this would be a hard problem.
https://xkcd.com/1425/ sort of applies.
should be called “squib game” as a squib is the offhand name for fake blood packs.
They kept it simple in the Korean version and shot people, but if MrBeast copied that exactly, it would be one way to have his video demonetised on youtube. And each of his videos makes so much from advertisements, he has a team of people from youtube helping him keep all his videos inside the rules and maximise everyones return on investment. Because alphabet always get most of the cake from the ads, while he and his team gets a reasonably good slice because his channel is in the top 5 most subscribed channels (93.4 million subscribers). That video alone has had nearly a quarter of a billion views in just four months. That is a lot of money to not be making if it was demonetised.
Some peepal MUST hear themselves talk…… forever.
Shoulda brought in Michael Reeves to also make the boards take people when they get shot as well.
Taze people I meant
Hehe I have been in the same project in the same setting. Even went to set up assembly line in china only to find that nothing was ready yet, even if assembly house claimed that where ready. On top of that they had not ordered PCBs yet and claimed 2 weeks leadtime on PCB in Shezhen.. lol.. (this was before covid)
Could have just used paintball guns on a motion sensing turret. Dual use for a Portal movie.
Paintball guns aren’t reliably accurate, or rather they weren’t when I was playing over a decade ago. Close range stuff sure go ahead but not at the distances seen in the video.
IMO the whole (very badly organized…) effort they put into this is completely lost in the resulting video. Even in the first game with all 456 contestants there is no confusion as to who has been removed. The exploding packs play zero role, they’re not even made to look like a cool effect and just seem to have been quite bothersome for everybody. Some participants commented handing out the packs took many hours and they worked so badly due to spotty WiFi that the whole thing had to be re-done a day later. You can see MrBeast removing people without even firing the packs in some games, e.g. having them jump into a foam-filled abyss instead.
TBH I actually don’t understand how the whole endeavour ended up costing an alleged 3.5 million USD. It’s not a particularly well-made video, even for MrBeast standards.
The way Williams friends work together as a engineering team in an emergency situation is exactly how I imagined beeing an engineer is when I was in middle school, fabulous!
Do you think William will open source the design of the igniter at some point?
If not does anyone know if there are some good DIY remote firework igniter designs out there?
The firmware and app are open source!
I dont get all the hate, this was a great video and a good look into the next generation of makers. Very cool to see all the cameos and how they all worked together. It’s very easy to just zero in on production and peak efficiency but sometimes ya just gotta have fun with an objectively terrible situation. I enjoyed every second of this vid. Also digikey back at it again with the components 3 bags deep.
It hurts to read so many bitter comments here. It is ok if you find the video to be too long, it is made to be entertaining for a general audience; it isn’t a tutorial or a lecture. The story is very relatable for anyone who’s done engineering projects with tight time constraints. He found an expensive way of doing it, an inexpensive version if the same that he wouldn’t be able to put together on time and ended up producing an alternative solution that worked OK and was inexpensive. He doesn’t deserve a Nobel prize but it is a cool story and I think it will inspire many young people who are considering engineering as a career. ain the era of AI, data and cryptobullshit we should be happy to see a youtube star who is into actual physical hardware like most of us.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)